Sentinel2 : get JPEG200 bands only

Question

I know it's possible to query products for Sentinel 2 via their SciHub, or using a tool like http://sentinelsat.readthedocs.io/en/master/

However, I need to somehow, given a product ID, get a URL to a specific band in JPEG200, like the following URL example:

https://scihub.copernicus.eu/apihub/odata/v1/Products('22e2fbfe-0aa7-423d-b0b5-df46527f03f5')/Nodes('S2A_MSIL1C_20170316T112101_N0204_R037_T29SNC_20170316T112658.SAFE')/Nodes('GRANULE')/Nodes('L1C_T29SNC_A009043_20170316T112658')/Nodes('IMG_DATA')/Nodes('T29SNC_20170316T112101_B04.jp2')/$value

This is the Band 4 of a given product.

Answer 1

Assuming you have the product identifier (22e2fbfe-0aa7-423d-b0b5-df46527f03f5) you can construct the query you want with a little parsing of the scihub odata response.

The constructed img_url can then be downloaded with the tool/library of your choice.

import requests
from six.moves.urllib.parse import urljoin

# connect to the api
api_session = requests.Session()
api_session.auth = ("username", "password")
api_url = "https://scihub.copernicus.eu/apihub/odata/v1/"

# product UUID you want to download a single band for
prod_id = "22e2fbfe-0aa7-423d-b0b5-df46527f03f5"

# parse the product name
nodes = api_session.get(urljoin(api_url, "Products('%s')/Nodes?$format=json" % prod_id)).json()
prod_name = nodes["d"]["results"][0]["Id"]

# parse the granule id
granules = api_session.get(urljoin(api_url, "Products('%s')/Nodes('%s')/Nodes('GRANULE')/Nodes?$format=json" % (prod_id, prod_name))).json()
granules["d"]["results"][0].keys()
gran_id = granules["d"]["results"][0]["Id"]


# parse the band names
bands = api_session.get(urljoin(api_url, "Products('%s')/Nodes('%s')/Nodes('GRANULE')/Nodes('%s')/Nodes('IMG_DATA')/Nodes?$format=json" % (prod_id, prod_name, gran_id))).json()
band_id = bands["d"]["results"][3]["Id"]  # element 3 is band 4, element 0 band 1

# construct the final image url
img_url = urljoin(api_url, "Products('%s')/Nodes('%s')/Nodes('GRANULE')/Nodes('%s')/Nodes('IMG_DATA')/Nodes('%s')/$value" % (prod_id, prod_name, gran_id, band_id))

Answer 2

I would recommend accessing data via the Sentinel-2 AWS Public Bucket.

Users can search for files using the index pages and access individual or related sets of files through the Sentinel Public bucket.

This service is designed for applications such as yours, where automated downloads are required. The bucket contains L1C (top of atmosphere reflectance) products like SciHub.

Answer 3

The following link may be useful: Download only some of the channels of a Sentinel-1 product